A new Grammarly feature called 'Expert Review' has been using the names of journalists and authors to lend credibility to its AI-generated writing suggestions, a practice discovered last week and confirmed to affect numerous individuals. The company, owned by Superhuman, did not seek permission from the people whose identities it employed, including prominent tech reporters.

In response to criticism, Grammarly has not apologized or removed the feature. Instead, it now offers an opt-out process. Those who wish to have their names removed must send a request to a specific email address. The company's statement, provided by Vice President of Product & Corporate Marketing Alex Gay, emphasized a desire to 'improve' the product and give experts 'greater control,' but made no mention of obtaining prior consent.

This solution places the burden entirely on individuals to discover the use of their name and act to stop it. For many, this discovery may never happen unless they or someone they know uses Grammarly's premium service. When pressed on whether it would implement more proactive protections, a company spokesperson said only that they are 'working on further refining the feature.'

The incident raises significant questions about digital identity and consent in an age of generative AI. Without a fundamental shift in policy, Grammarly's approach suggests a continued willingness to appropriate personal reputations by default, asking for forgiveness rather than permission.

Source: The Verge